Modeling of Geothermal Power System Equipped with Absorption Refrigeration and Solar Energy using Multilayer Perceptron Neural Network Optimized with Imperialist Competitive Algorithm

Abstract

For geothermal energy system, the low geothermal well operating lifetime and temperature is one of the main obstacles. Also, finding the optimum design parameters and operating conditions requires many experimental tests and intricate mathematical models. Besides, improving the energy efficiency and lacking technical feasibility of the combined geothermal systems are other challenges concerning geothermal systems. To cover the mentioned challenges, in the current study, a hybrid geothermal/absorption refrigeration system (ARS) incorporated with solar thermal collector, desalination unit and hydrogen storage system is designed and assessed. The proposed system is investigated by developing two methods of artificial intelligence (AI) as well as thermodynamic model. The intelligent methods are multilayer perceptron (MLP) neural network optimized with imperialist competitive algorithm (ICA), MLP-ICA, and MLP optimized with genetic algorithm (GA), MLP-GA. These methods are manufactured based on the solar irradiance, cooling water temperature difference, ambient temperature, pinch-point temperature, evaporating and condensing temperatures as independent parameters. These parameters are utilized to obtain the power generation, coefficient of performance of the ARS (COP chiller), heat exchanger area of the ARS, and cycle thermal efficiency. The obtained results show that simulation of the system by MLP-ICA was successfully carried out and this model operates substantially better than the MLP-GA for simulating the behavior of the system. Also, the payback time for the proposed system (with the interest rate of 3%) was obtained around 8 years.